Article
Computer Science, Artificial Intelligence
Abdelrahman Elsaid, Karl Ricanek, Zimeng Lyu, Alexander Ororbia, Travis Desell
Summary: Continuous Ant-based Topology Search (CANTS) is a novel nature-inspired neural architecture search algorithm based on ant colony optimization. It utilizes a continuous search space to automate the design of artificial neural networks, removing the limitation of predetermined structure sizes. By adding an extra dimension for neural synaptic weights, CANTS can optimize both architecture and weights, significantly reducing optimization time while maintaining competitive performance.
APPLIED SOFT COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Chao Liu, Lei Wu, Wensheng Xiao, Guangxin Li, Dengpan Xu, Jingjing Guo, Wentao Li
Summary: In this study, a novel variant of ant colony optimization algorithm called improved heuristic mechanism ACO (IHMACO) is proposed. It contains four improved mechanisms to enhance the efficiency and effectiveness of path planning. Experimental results show that IHMACO outperforms existing approaches in terms of path turn times.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Biology
Lei Liu, Dong Zhao, Fanhua Yu, Ali Asghar Heidari, Chengye Li, Jinsheng Ouyang, Huiling Chen, Majdi Mafarja, Hamza Turabieh, Jingye Pan
Summary: This study introduces a multilevel COVID-19 X-ray image segmentation method based on ant colony optimization. By improving the algorithm, it effectively enhances the diagnostic level.
COMPUTERS IN BIOLOGY AND MEDICINE
(2021)
Article
Computer Science, Artificial Intelligence
Yifeng Li, Ying Tan
Summary: In this paper, a theoretical model of fireworks algorithm based on search space partition is proposed, analyzed, and implemented. Experimental results show that the proposed algorithm outperforms previous variants of fireworks algorithm significantly, and achieves competitive results compared with state-of-the-art evolutionary algorithms.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Dickson Odhiambo Owuor, Thomas Runkler, Anne Laurent, Joseph Onderi Orero, Edmond Odhiambo Menya
Summary: Gradual pattern extraction is a field in Knowledge Discovery in Databases that aims to map correlations between attributes of a data set as gradual dependencies. In this study, three population-based optimization techniques are investigated to improve the efficiency of mining gradual patterns. The results show that ant colony optimization technique outperforms genetic algorithm and particle swarm optimization in the task of gradual pattern mining.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2021)
Article
Computer Science, Artificial Intelligence
Jafar Gholami, Farhad Mardukhi, Hossam M. Zawbaa
Summary: Meta-heuristic algorithms, such as the crow search algorithm (CSA), have shown promising results in solving optimization problems, but often suffer from issues such as local optima and premature convergence. This paper introduces an improved version, ICSA, which utilizes a new update mechanism to enhance convergence and local search ability. Experimental results demonstrate that ICSA outperforms traditional CSA and other meta-heuristic algorithms in terms of solution accuracy and efficiency.
Article
Computer Science, Artificial Intelligence
Ruyi Dong, Huiling Chen, Ali Asghar Heidari, Hamza Turabieh, Majdi Mafarja, Shengsheng Wang
Summary: The Kernel Search Optimization (KSO) algorithm was proposed to simplify the optimization process by transforming the optimization of nonlinear functions into a linear process. By adopting a local search of the hill-climbing algorithm and simplifying the calculation of kernel parameters, the improved algorithm outperformed the original KSO and some well-known algorithms in terms of accuracy and running time.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Information Systems
Yanfei Zhang, Yiyan Wen, Haiyang Tu
Summary: This paper proposes a shortest path planning method based on AIS data, which combines ACA and A* search algorithm to establish a high-precision environment model. The key points are extracted from the initial route obtained by the A* search algorithm, and the Bezier curve method is introduced to smooth the route to obtain the planned route. The experimental results validate the effectiveness of the proposed method in obtaining shorter paths faster and more efficiently.
Review
Automation & Control Systems
Huiling Chen, Chenyang Li, Majdi Mafarja, Ali Asghar Heidari, Yi Chen, Zhennao Cai
Summary: This paper provides a comprehensive review of critical studies related to the development of Slime Mould Algorithm (SMA), including an analysis of advanced versions of SMA and its application domains. The survey shows that SMA outperforms established metaheuristics in terms of speed and accuracy, and suggests possible future research directions.
INTERNATIONAL JOURNAL OF SYSTEMS SCIENCE
(2023)
Article
Engineering, Biomedical
Siti Khadijah Ali, Mohamad Faisal Fadzilan, Aida Nur Syafiqah Shaari, Muhamad Sukri Hadi, Rickey Ting Pek Eek, Intan Zaurah Mat Darus
Summary: Flexible beam structures are favored for being lighter and cost-effective compared to rigid structures, but are more susceptible to vibration. Researchers have developed various methods to suppress undesired vibration, with the cuckoo search algorithm (CSA) found to be the best model for representing the real behavior of flexible beam structures.
JOURNAL OF VIBROENGINEERING
(2021)
Article
Computer Science, Interdisciplinary Applications
Mahesh Kumar, Devender Kumar
Summary: The research utilizes the AMP variant of the ant colony optimization algorithm to overcome the problem of local optima in the gravitational search algorithm and improve the exploration ability of the search space. The proposed GSAMP algorithm is applied to fingerprint recognition for complete and latent fingerprints, showing superiority over other methods.
INTERNATIONAL JOURNAL OF MODERN PHYSICS C
(2023)
Article
Computer Science, Artificial Intelligence
Zhi-Gang Du, Jeng-Shyang Pan, Shu-Chuan Chu, Yi-Jui Chiu
Summary: In recent years, there has been rapid development in swarm intelligence, particularly with the emergence of the symbiotic organism search (SOS) algorithm. While the SOS algorithm has shown promise in solving real-life problems, its discretization poses challenges. This study introduces a multi-group discrete SOS algorithm with gene transfer and path cross strategies to enhance its performance in solving the TSP problem.
Article
Computer Science, Artificial Intelligence
Huijun Liu, Ao Lee, Wenshi Lee, Ping Guo
Summary: This study proposes a dynamic adaptive ACO (DAACO) algorithm to solve the TSP problem. DAACO introduces diversity in the initialization of the ACO algorithm by dynamically determining the number of ants to prevent local optimization. It also adopts a hybrid local selection strategy to improve the quality of optimization and reduce the time required. Experimental results demonstrate that the DAACO algorithm outperforms existing state-of-the-art ACO algorithms in terms of convergence time, solution quality, and average value on the TSPLIB dataset.
COMPLEX & INTELLIGENT SYSTEMS
(2023)
Article
Mathematical & Computational Biology
Xiao Yang, Xiaojia Ye, Dong Zhao, Ali Asghar Heidari, Zhangze Xu, Huiling Chen, Yangyang Li
Summary: This study proposes a new multi-threshold image segmentation model based on the two-dimensional histogram approach, using an enhanced ant colony optimization algorithm combined with two-dimensional Kapur's entropy to search for optimal thresholds. Experimental results demonstrate that the proposed model outperforms the comparison method in segmenting images and provides high-quality samples for subsequent analysis of melanoma pathology images.
FRONTIERS IN NEUROINFORMATICS
(2022)
Article
Computer Science, Theory & Methods
Joshua Peake, Martyn Amos, Nicholas Costen, Giovanni Masala, Huw Lloyd
Summary: This paper presents an improved algorithm for the Virtual Machine Placement (VMP) problem, which significantly improves the solution speed by utilizing parallelization techniques and modern processor technologies. The algorithm achieves solution qualities comparable to or even superior to other nature-inspired methods.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2022)
Article
Geosciences, Multidisciplinary
Yongliang Chen, Wei Wu
NATURAL RESOURCES RESEARCH
(2019)
Article
Geochemistry & Geophysics
Yongliang Chen, Wei Wu, Qingying Zhao
Article
Geosciences, Multidisciplinary
Yongliang Chen, Wei Wu, Qingying Zhao
NATURAL RESOURCES RESEARCH
(2020)
Article
Geosciences, Multidisciplinary
Yongliang Chen, Shicheng Wang, Qingying Zhao, Guosheng Sun
Summary: Isolation forest and elliptic envelope models were used to detect geochemical anomalies, with the bat algorithm optimizing the parameters. The bat-optimized models showed improved performance in geochemical anomaly detection, with the optimal threshold determined by the Youden index. Compared to anomalies detected by the elliptic envelope models, anomalies detected by the isolation forest models exhibited higher spatial relationship with mineral occurrences.
JOURNAL OF EARTH SCIENCE
(2021)
Article
Computer Science, Interdisciplinary Applications
Yongliang Chen, Guosheng Sun, Qingying Zhao
Summary: The study utilized eight Distance Anomaly Factors (DAFs) to map gold potential in Inner Mongolia, China, showcasing their effectiveness in comparison to the One-Class Support Vector Machine (OCSVM) for gold potential mapping. The optimal threshold for distinguishing gold potential cells was determined by maximizing the Youden index, with the targets predicted by the DAFs occupying 7.4% - 16.5% of the study area and containing the majority of discovered gold deposits.
EARTH SCIENCE INFORMATICS
(2021)
Article
Geology
Yongliang Chen, Qingying Zhao
Summary: The recursive indicator elimination (RIE) method, which combines recursive elimination process and machine learning techniques, is proposed to determine the optimal subset of geochemical elements for mineral exploration targeting. The method showed the best performance in a case study, with results consistent with geological characteristics.
ORE GEOLOGY REVIEWS
(2021)
Article
Geology
Yongliang Chen, Yanhui Sui
Summary: This study introduces the use of dictionary learning techniques for mineral prospectivity modeling, which outperforms logistic regression and one-class support vector machine models and shows strong consistency with geological and metallogenic characteristics in the study area. The high-performance of the dictionary learning algorithms suggests their potential for further exploration targeting of different mineral deposit types in various areas.
ORE GEOLOGY REVIEWS
(2022)
Article
Computer Science, Interdisciplinary Applications
Yongliang Chen, Yuanqing Zhang, Yulei Tan
Summary: The study extended the prospecting cost-benefit strategy to use MCC and F-measure to represent mineral potential, and compared it with the bagging and boosting ensembles in polymetallic prospectivity modeling. Results showed that the prospecting cost-benefit strategy outperformed ensemble learning models in polymetallic prospectivity modeling.
EARTH SCIENCE INFORMATICS
(2022)
Article
Geosciences, Multidisciplinary
Yongliang Chen, Chenyi Zheng, Guosheng Sun
Summary: In this paper, a methodology for gold prospectivity modeling was developed by combining Laplacian eigenmaps (LEMS) and the least angle regression (LARS) in the Jinchanggouliang area, an important gold metallogenic district in China. Performance evaluation revealed that the LEMS-LARS model, the LGR model, and the OCSVM model yielded comparable performances and outperformed the LARS model for gold prospectivity modeling in the study area. The optimal gold prospective areas delineated by the LEMS-LARS model, the LGR model, and the OCSVM model were spatially associated with areas having favorable metallogenic conditions.
NATURAL RESOURCES RESEARCH
(2022)
Article
Geochemistry & Geophysics
Yongliang Chen, Alina Shayilan
Summary: In geochemical exploration, the performance of anomaly detection models is influenced by the lack of consideration for the relationship between geochemical elements and mineralization. This study combines neighbourhood component analysis and dictionary learning algorithms to improve the performance of dictionary learning models in geochemical anomaly detection.
GEOCHEMISTRY-EXPLORATION ENVIRONMENT ANALYSIS
(2022)
Article
Geology
Yongliang Chen, Yanhui Sui, Alina Shayilan
Summary: Both anomaly detection algorithms and supervised classification algorithms can be used to detect mineralization-related geochemical anomalies in areas with discovered mineral deposits. However, neither of these models perform well due to their limitations in utilizing known mineral deposits as supervisors or handling extreme class-imbalance in geochemical exploration data. Therefore, a self-training model based on support vector classifiers was adopted and outperformed other models in detecting gold mineralization-related geochemical anomalies.
ORE GEOLOGY REVIEWS
(2023)
Article
Geosciences, Multidisciplinary
Yongliang Chen, Laijun Lu
Summary: Unsupervised anomaly detection techniques model the distribution of geochemical exploration data, while supervised classification techniques make use of mineral deposit information to distinguish anomalies. However, the data imbalance in distinguishing anomalies from the background poses a challenge.
MATHEMATICAL GEOSCIENCES
(2023)
Article
Geology
Yongliang Chen, Xudong Du, Min Guo
Summary: The self-paced ensemble algorithm is a more efficient and robust approach for detecting mineralization anomalies in geochemical exploration data compared to the self-training algorithm. Through a case study in Inner Mongolia, it is shown that the self-paced ensemble algorithm outperforms the self-training algorithm in terms of classification performance, robustness, and efficiency.
ORE GEOLOGY REVIEWS
(2023)
Article
Geology
Jiaxing Chen, Yongliang Chen
Summary: Accurate identification of mineralization anomalies is crucial in geochemical exploration. This study utilizes graph convolutional extreme learning machines (GCELMs) and voting-based GCELM (V-GCELM) model to identify mineralization anomalies, and the results show that V-GCELM model performs the best. These techniques show promise for building high-performance classification models for identifying mineralization anomalies from geochemical exploration data.
ORE GEOLOGY REVIEWS
(2023)
Article
Geochemistry & Geophysics
Om Prakash Kumar, Amiya S. Naik, P. Gopinathan, T. Subramani, Vishvajeet Singh, Prakash K. Singh, Uma K. Shukla, Arun Prabhu
Summary: This study characterizes lignite samples from Kapurdi, Giral, and Sonari mines in Rajasthan's Barmer Basin using petrographic and geochemical techniques. The results provide insights into the geochemical properties, hydrocarbon potential, depositional environment, and paleo-climatic conditions of these lignite deposits. The study finds high volatile matter and sulfur concentrations in the Barmer lignite deposits, as well as indications of a wet environment during organic material decomposition. The findings have implications for understanding the coalification profile and hydrocarbon source rock potential in the region.
JOURNAL OF GEOCHEMICAL EXPLORATION
(2024)
Review
Geochemistry & Geophysics
Ashkan Jahandari, Behnam Abbasnejad
Summary: The assessment of heavy metal concentrations in agricultural soil is important for crop safety and quality, as well as potential risks to human health. This review analyzed published data on seven heavy metals in Iranian farmland soils and found that nickel and cadmium exceeded permissible levels. Arsenic showed significant contamination in multiple provinces, while lead and cadmium were highly contaminated in Fars province. Overall, the contamination levels of heavy metals in Iranian agricultural soils were relatively low.
JOURNAL OF GEOCHEMICAL EXPLORATION
(2024)
Article
Geochemistry & Geophysics
Imran Ud Din, Said Muhammad, Shah Faisal, Inayat ur Rehman, Wajid Ali
Summary: The presence of heavy metal contamination in coal mines and surrounding environments in the Hangu and Kurram districts poses potential risks to the environment and human health.
JOURNAL OF GEOCHEMICAL EXPLORATION
(2024)
Article
Geochemistry & Geophysics
Douglas Almeida Silveira, Paola Ferreira Barbosa, Cassiano Costa e Castro, Guilherme Ferreira da Silva, Joseneusa Brilhante Rodrigues
Summary: This study analyzed apatite grains from the Angico dos Dias Carbonatite Complex region in Brazil using an electron probe microanalyzer. The results showed that the apatite grains had characteristics of metacarbonatitic lithotypes and some had undergone substitution processes. Statistical analysis helped identify the sources of the apatite grains and confirmed the presence of carbonatitic bodies in the area.
JOURNAL OF GEOCHEMICAL EXPLORATION
(2024)